Treatment of Idiopathic Aplastic Anemia (AA) with Expanded Autologous Regulatory T Cells: The Tiara Phase One Trial

Introduction

Regulatory T cells (Tregs) are reduced in both severe and non-severe AA. We have shown that Tregs in AA show high expression of CD45RA and CD27 characteristic of a naïve/resting phenotype and expand in vitro in the presence of IL-2 (1000IU/mL). Expanded Tregs are ≥90% FOXP3⁺, suppressive, stable, polyclonal and prevent graft vs host disease in mouse models. These observations laid the foundation for the first in-man Phase 1 clinical trial of expanded autologous Tregs in AA (EudraCT Number: 2021-000082-33)

Methods

Trial entry criteria included patients ineligible for Hematopoietic Stem Cell Transplantation or failed/ ineligible for Anti thymocyte globulin + cyclosporine A/ Eltrombopag. The trial design includes 2 doses of expanded Tregs (5×10⁶ /kg) administered 2 weeks apart. 2.2 – 2.9 (avg 2.6) times total body volume was processed to achieve an apheresis volume of 120-200mL. Tregs were isolated from the apheresis product and expanded between 16-23 days to achieve the required cell dose.

Deep cytometry, single cell sequencing [Antibody sequencing (AbSeq), Single-cell RNA (scRNA) and Single-cell TCR (scTCR) sequencing] and cytokine profiling was performed on blood samples collected prior to infusion on day 1 and 15 and post infusion of Tregs on days 29, 58 and 180.

Results

Tregs were successfully expanded from all 6 patients (severe=3, non-severe=2, and very severe =1) from a starting number of median (92.43×10⁶ ) cells. Tregs were expanded in sufficient numbers (2.5-6.8 ×10⁹cells) and met the release criteria including viability (≥70%), phenotype (≥30% CD4⁺CD25⁺CD127, low FoxP3⁺), impurities (≤10% CD8⁺ cells), bead concentration (≤ 100 beads per 3×10⁶ cells), sterility, endotoxin negativity and potency (≥60% suppression of Teff proliferation).

Six patients median age 74 (68-79 yrs) completed the treatment with both doses without any safety concerns, except 1 patient who developed antibiotic-responsive fever (38°C) within 24 hrs of the 1^st infusion but received the 2^nd dose without any side effects. Manufacturing process failed in 1 patient, but cells were successfully expanded at the 2^nd attempt in this patient.

One transfusion-dependent patient became transfusion-independent, with platelet count increasing from 10×10⁹/L before the 1st infusion to 35×10⁹/L at 12 months. Another patient also showed improvements with platelet count increasing from 77×10⁹/L to 96×10⁹/L at Day 263 following the 2^nd infusion. The patient achieved red cell transfusion independence lasting 100 day⁺.

Mass cytometry on infusion products and longitudinal samples from 5 of the 6 patients identified distinct clusters of regulatory T cells. Some clusters were present at all time points (pre-existing Tregs), while others appeared only post-infusion (post-treatment Tregs). Post-treatment Tregs were characterized as FoxP3dimCD25⁺CD95⁺CCR4⁺IL7R^-CD39^- with variable expression of CCR6 and CD161. The frequency of pre-existing Tregs did not change during treatment, whereas maximal expansion of post-treatment Tregs occurred at Day 29 post-1^st infusion, ranging from 5.01% - 30.7% of total CD4⁺ T cells, sustained until Day 58 (4.3% - 18.5%).

Responders (n=2) showed higher frequencies of post-treatment Tregs than non-responders (n=3), both at Day 29 mean (34.55% vs 6.4%) and at Day 58 (15.1% vs. 5.15%). scRNA sequencing in 1 responder confirmed the accumulation of distinct post-treatment regulatory T cell subsets. scTCR sequencing showed that post-treatment Tregs had greater TCRVβ repertoire overlap with the infusion product (10-24%) compared to pre-existing Tregs (5%), suggesting in vitro-expansion of rare regulatory T cell clones. Infusion products from different patients varied in the number and expression of Ki67 and PD1. The highest numbers of Treg-like cells were observed in patients with high Ki67⁺ PD1^- Tregs numbers in the infusion product (67% vs 19%-27% in other patients) suggesting that less differentiated Tregs might result in better responses.

Conclusion

Our results show that Tregs from AA patients are expandable, safe to infuse and may dampen the abnormal immune milieu associated with AA. In addition, we have been able to track the persistence of the expanded Tregs in vivo, correlate clinical efficacy with the accumulation of clusters of regulatory T cells post-infusion, and identified phenotypic markers in the infusion product that correlate with in vivo expansion.